AI red-teaming framework for testing LLMs and generative AI systems for jailbreaks, prompt injection, harmful content, data leakage, and multi-turn attacks. USE WHEN user mentions AI red team, LLM jailbreak, prompt injection test, AI risk assessment, AI security testing, GenAI red teaming, NIST AI RMF MEASURE, ISO 42001 testing evidence, or wants to stress-test an AI system. Wraps Microsoft's open-source PyRIT (Python Risk Identification Tool) v0.13+.
